2.0 Analysis 2.1 General The pilots of both aircraft were certified and qualified, and there was no evidence that any physiological factors affected their ability to conduct the flights safely. There were no mechanical discrepancies found with either aircraft that would have contributed to the occurrence. The only apparent evasive action to avoid the collision was taken by the Air Sandy pilot; however, it was taken in insufficient time to avoid the collision. Although this was a typical VFR situation in that two aircraft were flying under visual conditions, a number of factors combined to create a high risk of collision. Therefore, this analysis will examine the pilots' ability to see the other aircraft, the limitations of the see-and-avoid concept, the lateral precision of GPS, why the pilots were not alerted, the effects of high closing speeds, the apparent lack of pilot understanding of how to optimize avoidance manoeuvring, and other factors that could have contributed to the occurrence. 2.2 Limitations of the See-and-Avoid Concept See-and-avoid is used as the primary means of separating aircraft in visual flight conditions; however, due to the physiological limitations on the human visual and motor-response systems, it may be impractical to rely on this system as the primary means of separation. This is particulary true in situations that involve head-on geometry with high closing speeds; the risks involved in relying on see- and-avoid increase as the relative closing speed of the aircraft increases. In this occurrence, the Bearskin Metro had been descending at approximately 250 knots and was decelerating to approximately 230 knots. The departing Navajo would have just levelled off from an en route climb and would have been accelerating to a flight planned cruise speed of approximately 180 knots. At a closing speed of about 410 knots and with 12 seconds before impact, the Bearskin pilots had less than a 16% probability of detecting the Air Sandy aircraft, and the Air Sandy pilot had less than a 13% chance of detecting the Bearskin aircraft. Not until the Bearskin pilots were within 4.4 seconds of impact and the Air Sandy pilot was within 3.5 seconds of impact did the pilots of either aircraft have at least a 50% chance of detecting the other aircraft. It is likely that the Air Sandy pilot saw the Metro and attempted an evasive manoeuvre. With 12 seconds required to see and avoid another aircraft, it is doubtful that a pilot could effectively avoid another aircraft on a head-on collision course at a high closure rate. Advisory communications by Air Traffic Services and flight crew monitoring of frequencies are crucial in helping prevent collisions. It is, therefore, considered important that Flight Service specialists have a level of understanding comparable to that of pilots on the limitations of see-and-avoid. 2.3 Avoidance Manoeuvre Based on the aircrafts' relative attitudes at impact, it appears that the Air Sandy pilot had detected the approaching Bearskin Metro, and had begun an evasive manoeuvre by initiating a steep left-turn which had reached approximately 45 degrees to 60 degrees of bank. As the bank angle increased, the aircraft's cross-section would have increased correspondingly from a minimum value of approximately 13 feet to some final value in the range of 28 to 34 feet; the net result of the evasive roll would have been to inadvertently increase the Navajo's cross-sectional area and, thus, increase the risk of colliding with the oncoming Metro aircraft. A vertical manoeuvre, consistent with that demanded by TCAS-equipped aircraft, is normally more effective in close-range, head-on collision scenarios. However, formalized training on how to recognize in-flight collision geometry and on how to optimize avoidance manoeuvring is not part of the required syllabus for any level of civilian pilot licence in Canada. Without appropriate training, it is possible that a pilot who sees a target in sufficient time to react may react by turning, thereby increasing the cross-sectional area of the aircraft and increasing the risk of collision. 2.4 FSS Advisory Communications In this occurrence, it is unlikely that the involved aircraft were on the same radio frequency at the point where the collision occurred. Since the collision occurred approximately 12 nm from the airport, outside of the MF zone, it is likely that the Air Sandy aircraft would have been on 122.8 MHz and would not have heard any of the radio transmissions between the FSS and the Bearskin aircraft. The inbound Bearskin flight was operating under IFR control and had been directed to change from Winnipeg ACC to the Sioux Lookout MF. It is therefore unlikely that either aircraft would have heard transmissions generated by the conflicting flight. The only common frequency for the two aircraft would have been the MF, but, when the collision occurred, both aircraft were well outside the 5 nm control zone. If the flight crew of either aircraft had been alerted to the presence of the other, the likelihood of seeing the other aircraft would have increased by about a factor of eight. This increased likelihood of detection might have given the crew of either aircraft the opportunity to initiate a collision avoidance manoeuvre in time to prevent the collision. Unlike air traffic controllers, Flight Service specialists do not have radar equipment. The specialist at Sioux Lookout was provided with VDF equipment, which has some capability to display potential collision information; however, VDF does not provide distance information. The specialist had to rely on voice reports from aircraft, and, aided by a plotting board, his own ability to keep mental track of traffic in the area to visualize potential conflicts. He did not have additional equipment that would have helped alert him to the presence of the Bearskin flight and the conflict between it and the Air Sandy flight. 2.5 Control Zone The use of the MF is only required while in the control zone. The edge of the control zone in Sioux Lookout is 5 nm from the airport. High performance, multi-engine aircraft that frequent the airport can have closing speeds of up to 400 knots (one departing, one landing), which is equivalent to more than 6 nm per minute. If the control zone radius is 5 nm, it is likely that these aircraft will have less than one minute before their paths cross within the control zone. At such a speed, there is little time for the Flight Service specialist to convey traffic information, especially if he is tasked with other duties such as talking to other traffic or to the ACC on the telephone. If the MF area were larger, or if the aircraft were approaching and departing at lower speeds, there would be more time for aircraft to be made aware of each other. 2.6 Lateral Precision of GPS